Dennis Hackethal’s Blog
My blog about philosophy, coding, and anything else that interests me.
Tweets
An archive of my tweets and retweets through . They may be formatted slightly differently than on Twitter. API access has since gotten prohibitively expensive – I don't know whether or when I'll be able to update this archive.
But in case I will, you can subscribe via RSS – without a Twitter account. Rationale
Wenn dich AI interessiert, hör dir das mal an. Hoffe, dir geht’s bald besser. cbc.ca/amp/1.4696754
It's a never ending, beautiful journey. twitter.com/Calvinn_Hobbes…
Also nonsense. Our minds are much better than any other animal's because we can create new explanatory knowledge. Innovations help us prosper. Progress is achievable and has been achieved only thanks to reason and science. Advocating against using "too much" reason is anti-human.
Nonsense. People are universal explainers, unlimited progress is possible. Read "The Beginning of Infinity" by David Deutsch.
- What’s the issue with logic, reason and science? Name one case where using them less would be better.
- We don’t put faith in them. That would itself be illogical, unreasonable and unscientific. Faith means “don’t ask questions”.
RT @ToKTeacher:
Justified True Belief. https://t.co/e8IdnlIVIL
Is this a little bit like life? And if so, would understanding the origin of life help us understand how creativity works, and vice versa?
2/2
The creative algorithm was originally used to replicate memes. When turned on itself, it tries to replicate creativity (e.g. in AGI research); it tries to replicate itself.
1/2
As in “falsifiable by observation”? It can’t be. It’s a philosophical principle. You can criticize it, though.
@ChipkinLogan @DavidDeutschOxf
We should expect our theories to contain mistakes. Problems with constructor theory may sooner or later be discovered, which will require an even better/deeper explanation. There are no “final” theories.
@dela3499 @DavidDeutschOxf @dautingthomas
Yeah I wonder what an example would be as well. Up until now I thought there was only one universal creativity. Is it that there would be a difference in code quality across different implementations of creativity?
@dela3499 @cthulhupotamus @DavidDeutschOxf @bensomer_ville @fieryfalliblist @Crit_Rat @gleason_colum
Love me some Debussy Delight. I also recommend Stravinsky Strudel.
As an individual’s knowledge grows, so does the amount of information he can use to create new explanations. Therefore, I would expect his creation of new knowledge to get slower over time. Yet we don’t seem to observe that. How can we explain that?
Oooh I love listening to new material of his. Thanks for sharing!
None? My phone can recognize me, too. I don’t think it’s conscious.
Damn, that’s cray cray in a good way right there!
I still don’t get it. Why do we need an alternative theory to falsify another? If I have a theory and any one of its predictions doesn’t come true, the entire theory is refuted by modus tollens alone, no?
Can you please elaborate on “Theories don’t need to be falsifiable to be science.” They need to be hard to vary, but also falsifiable, no?
RT @ChipkinLogan:
The vast majority of the history of humanity is mired with tyranny, poverty, and misery.
How lucky we are to live in the…
RT @ChipkinLogan:
Explanation, not prediction, is the purpose of science @bgreene twitter.com/bgreene/status…
Regrettably, yes. The three stages of adopting critical rationalism that I’ve observed: 1) “Anything that’s not induction is stupid or navel gazing” 2) “Actually, this is neat!” moves on to something else 3) “Holy shit this is amazing and I’ll apply it” Few make it to 3 :(
Agree with Brett here. I’ve read four or five of Popper’s books and still struggle with LScD. My first Popper book was “All Life is Problem Solving”, after I read BoI and FoR (in that order). I think it’s the last book he published before his death. It’s great!
RT @DoqxaScott:
"Everyone can and should be a scientist. Because being a scientist just means wanting to understand the world, and using th…
@ToKTeacher @mizroba @LTF_01 @Locus_of_Ctrl
Not to mince words, but I wonder if we should distinguish self awareness from self recognition? Self recognition is the thing we could easily program today, whereas self awareness requires creativity.
@TobyJIB @DavidDeutschOxf @dela3499 @ChipkinLogan
Given what the announcer said moments before, it seems to me he’s just restating the final anthropic principle. But I can’t say for sure.
A friend from college and I would watch many episodes of South Park and a few animes (Death Note, Elfenlied, and others) that way. We’d talk on Skype and play the video at the same time and he would usually mute his video while listenting to my sound.
That’s not true. You don’t know what to observe without some hypothesis of what to look for and where something of interest might be.
I don’t disagree though that we should always try to falsify theories and demand good corroborating evidence, the absence of which leaves a theory vacuous.
I’d agree here with @jeffreyketland’s remark. Also I’d argue that science begins with problems and the quest for good explanations.
Eg Newtonian physics is still very useful and explains many things in certain contexts where relativity would be overkill to use, even though relativity is of course the better explanation. 2/2
I haven’t read Lakatos, but they may indeed uphold even a falsified theory if it has good explanatory power. 1/2
@JulieKuhrt @ChrHansen @cobalt_io @tijuanera @Recruit_and_U @CarolineWMWong @shark_256 @MrJakobStorm
Wow @ChrHansen your team has gotten so big! Congratulations!!
I like the emphasis on conjectures and refutations. Yet sometimes in interviews you say that we should consider something false until proven true. How do you reconcile these two approaches? (The difference being that conjectures in the Popperian sense can never be proven true.)
Are you working on a new book? Please say yes. :)
@cloroussy @DavidDeutschOxf @Quillette
Great article, thanks for sharing!
RT @royalsociety:
Robert Boyle FRS was born #onthisday in 1627. A founding member of the Society, he laid the foundations of #chemistry and…
RT @ToKTeacher:
Some excerpts of "Artificial Creativity" (Chapter 7 of "The Beginning of Infinity") by @DavidDeutschOxf with some remarks h…
Except it isn’t here yet at all. Hyperbolic statements like these impede research. aeon.co/essays/how-clo…
There’s also no difference between the sexes, and the sooner men realize this the better. ;-)
Yes, I’ve been practicing “getting it over with” because of the same reasons. It’s had a very positive impact on my life even when I do it just with small things such as “do the dishes” or “I’m already lying in bed but should brush my teeth now before I get more tired”.
RT @NatGeo:
"Since [Long-tongued bats] are active in complete darkness, photography is a challenge," writes photographer Sigal Cohen, who g…
RT @Crit_Rat:
The world has never been more free from suffering than it is now, and if we want to, we can gradually expunge all the remaini…
RT @DavidDeutschOxf:
I was a member of an audience of schoolchildren to whom he said this in 1971.
I think few of us believed it then. But…
You might enjoy David Deutsch’s “The Beginning of Infinity” about how special humans are. I’d love to see him on your show.
RT @DavidDeutschOxf:
Sustainability is unsustainable medium.com/@jrleahenry/su…
@wuweiwave @reasonisfun
"[Euclid] put into his book only those problems that he could solve. And this, in a certain sense, was a crime."
I LOVE that :)
So would you say machine learning is orthogonal to AGI? Or does it contribute in some way?
@mattguttmanorg @DavidDeutschOxf
I don't think that's true. @DavidDeutschOxf provides a good explanation of why behaviorism is not a good epistemic frame. See aeon.co/essays/how-clo… Don't mean to put words in his mouth though if I'm wrong in this particular context!
5/5 And each time this is done through conjecture and criticism. And then the cycle continues. Does this not seem Popperian? If it is, it seems to me we have made at least some headway in trying to code up AGI, but we'd need to do it for more than just mathematical problems.
4/5 In this example, the first problem is "We don't have an explanation for how the planets move", to which the conjectured explanation is Kepler's laws. The second problem is the observation that they're not accurate enough, and so the laws are improved, etc.
3/5 Then, if there is a problematic observation (problematic = it cannot be derived from Kepler's laws, such as a planetary position that does not fit the model), then the symbolic regression runs again given the problem, and it may find better explanations than Kepler's.
2/5 What I mean is, eg, symbolic regression can find Kepler's laws given data about planetary positions etc. And it does so through conjecture and criticism, since it's a genetic algorithm. I think John Koza may actually have done this for Kepler's third law, but I'm not sure.
1/5 We are on the same page about science being a process of conjecture and criticism that starts with problems.
I recently stumbled upon the concept symbolic regressions. Seems related to Popperian epistemology and explanations. Is being a universal explainer the same as being able to construct universal symbolic regressions (as opposed to just mathematical ones)?
It says the video is unavailable (for me at least).
What's your take on machine learning? It strikes me as induction applied to computer engineering and if I had to guess it does not bring us one step closer to general AI, but I'm curious if that's false.
@ToKTeacher @CoatOfPainter
I love that chapter in Beginning of Infinity!
@kpschoedel @reasonisfun @ToKTeacher @dela3499
"Notorious KRP"!
@RJsnda @ToKTeacher @DavidDeutschOxf
In the "Logic of Scientific Discovery" chapter 1 section 2, Popper conjectures that intuition plays an important role in coming up with new explanations (creativity). He also quotes Einstein as corroborating this in his address at Planck's 60th birthday.
@Clorwall @DavidDeutschOxf
Yes I loved that one! Perhaps in that same vein, “adapted information” is another one?
RT @DavidDeutschOxf:
That link doesn't seem to work for some people. This one works: stitcher.com/podcast/the-te… twitter.com/TEDchris/statu…
RT @jmsbutcher:
REPLICATING MEMES was the original genetic purpose of our ability to create knowledge, explain the world, and basically eve…
Instead, he describes the process of conjecture and criticism by which theories can be dismissed. Surviving theories are nonetheless unlikely to be true. That likelihood does not improve with more failed tests. (See Logic of Scientific Discovery, Popper, 1934) 2/2
Loved your book "The Emperor of All Maladies" - just one minor suggestion for improvement. Popper does not describe the process by which scientists "verify" theories. And theories do not gain credibility or robustness when attempts to falsify them fail. 1/2
Excellent article about what's standing between us and AI: aeon.co/essays/how-clo…
It is my understanding that to apply knowledge, one has to have knowledge that said knowledge exists in one’s mind. But one has to have knowledge of that second knowledge too, etc. An infinite regress?
RT @elonmusk:
@SamHarrisOrg @SpaceX
We’re doing ok for a bunch of monkeys. Humanity rocks!
Interesting, just the other day I started familiarizing myself with Hannah Arendt and her concept of the “banality of evil”. Can you elaborate where you find fault with it?
RT @DavidDeutschOxf:
Interesting, if so, that would seem to violate the principle of interoperability of information, as proposed in constr…
RT @cardboardisaac:
Fantastic list! And nearly all of them seem to have been well accomplished. twitter.com/royalsociety/s…
RT @MarsCuriosity:
I'm back! Did you miss me? This selfie is part of a fresh batch of images, direct from #Mars. Check out all my raw image…
A great example of what can happen when you look for confirmation, not refutation: youtu.be/bozH0O_cVhY
Unfortunately no subtitles, but brilliant nonetheless. youtu.be/1467dwUh5Nw
RT @ClickHole:
This Plus-Size Model Was Inspiring. But Then She Lost 100 Pounds, Which Was… Also Inspiring? Even Though She Was Already Per…
RT @MarsCuriosity:
Meanwhile, back on Mars... I’m checking out these stick-like figures. Each is about a quarter-inch long. Maybe they're c…
RT @ClickHole:
Awesome: When A Little Girl Told Neil DeGrasse Tyson She Wanted To Live On Jupiter, He Completely Shut Her Down https://t.co…
RT @ClickHole:
Embarrassing: The U.S. Is Ranked 182nd In The World Alphabetically clckhl.co/uCQALuX https://t.co/9Jk8uBErFh
RT @ClickHole:
Awesome: When This 8-Year-Old Girl Told J.K. Rowling That She Liked ‘Harry Potter,’ The Author Said ‘Yeah, No Shit’ https://…